I record in MXV then export in MPEG2. Funny I tried exporting in MPEG4 but the quality is considerably worse.

There are no AV inputs on my camera, unfortunately.

My VCR does not have an S-Video output socket. I just used a SCART adapter with an S-Video output cable attached, Would an RCA - S-Video cable help?

I want to avoid the expense of digitising my VHS tapes commercially. I'd be happy with the quality produced by Magix Video Easy, if it were not for the interference.

My VCR is a Daewoo QB12PI. My camera is an old Sony Handycam Vision CCD-TRV78E

By the way, there is little or no wobble in my VHS captures. Should TBC eliminate the banding/flicker/interference as well?

It's a wonder Magix didn't do something to address this problem (capture card/software) as I'm sure it must be a common complaint.

@cmclernon

My VCR does not have an S-Video output socket. I just used a SCART adapter with an S-Video output cable attached, Would an RCA - S-Video cable help?

I don't think a Composite to S-video cable will work because the luma (brightness/contrast) and Chroma (colour) signals would have to be split by the cable.

My VCR is a Daewoo QB12PI

Thanks, I can't find any info on that. If it doesn't say it's an S-VHS machine, it probably won't have S-Video Out.

By the way, there is little or no wobble in my VHS captures. Should TBC eliminate the banding/flicker/interference as well?

Nothing's guaranteed in this business. but I'm pretty sure putting some sort of TBC device eg a ES-10, ES-15 or Sony HXD-890 or a Panny DVD/VHS combo such as a ES-35 or EZ-48 (which both have built-in stabilizers, as per the ES-10 and 15) will correct all the issues you're having.

The whole situation is annoying because there are so many variables in recording quality, VCR quality (I have a newish cheapy that produces better a better picture than a 3x the price S-VHS VCR) and then the capture device. Even the capture software may affect the outcome.

It's a wonder Magix didn't do something to address this problem (capture card/software) as I'm sure it must be a common complaint.

I don't think it's a Magix problem, per se, assuming their videowandler is reasonable quality. I have 5 capture sticks here and they all basically produce the same result.

I record in MXV then export in MPEG2. Funny I tried exporting in MPEG4 but the quality is considerably worse.

That's good. Set the highest quality MXV that you can for capture. Your MP4s should look as good as the MXV; make sure you have the export bit rate nice and high. For Standard Def (720x576 or 480) use around 6,000kbps if you have the ability to set it for export.

@AAProds

Hi Al.

Please forgive me for interrupting but you have piqued my curiosity on this subject now as I am all about trying to preserve quality of any transfer or if possible even enhancing it. So I have a few questions.

The use of a TBC. Does it specifically deal with the cross hatch moving pattern we see in this video or does it do more?

Is its job to line up the top bottom fields more precisely or have something to do with timing or both?

If you think this is a necessary step then it looks like looking for a better machine is a must do first step or all else could be a waste of time.

Do you think RYVT will have enough tools to help clean up the files further if needed after that purchase?

I don't know if RYVT has the same ability to produce as high a quality MXV file as MEP.

@cmclernon uploaded the MP4 file version which they have stated was the poorer quality of the export formats.

Would it be possible to view the MXV file in MEP to see if the file size would be comparable to the MXV file sizes MEP can produce?

Would there be any advantage of recording the file to AVI? Is it possible in RYVT?

I noticed where there is a relatively clean frame in the sample file that the MP4 has blocked whole areas of colour. Possibly due to the lack of a colour pallet from the VHS medium.

That is what is concerning me. Where there is not much detail or lack of interference patterns, the colour areas are being rendered in (relatively) large blocks. I think some sharpening or contrast adjustment at that stage may help if the mxv file has enough resolution or am I wrong? Also the exported final files needs to have less compression applied to my mind.

Would that be possible with RYVT?

Has MEP the ability to capture an analog video stream any better? Would it be worth the upgrade?

I don't know yet whether cmcleron's goal is to put these files onto DVD / Blu-ray or to keep them as Digital computer files for better quality playback.

I would really like to see the mxv file uploaded to see if it is clean enough.

I know that RYVT has a much lower machine specification to run than MEP and that we have not asked

cmcleron for their machine specification yet or whether they want to pursue this option.

@cmclernon

Hi.

It seems to me that the clip in question at least is of a reasonable quality and could be improved on if you follow Al's advice.

If you could upload your mxv file from this clip to see if it is just suffering from a lack of a TBC or not may be of some benefit if it reduces most of the problems you are having with exporting quality.

It could well be that if the file has enough data in it that recording the same off of a better VCR and then exporting the file at a much higher bit rate may be enough without having to go through more steps to clean up the image if some of my assumptions are correct above and Al verifies my conclusions.

I may be completely wrong as it is not my main area of experience.

As I said at the beginning, I'm just curious.

Ray.

@CubeAce @cmclernon

The use of a TBC. Does it specifically deal with the cross hatch moving pattern we see in this video or does it do more?

Is its job to line up the top bottom fields more precisely or have something to do with timing or both?

If you think this is a necessary step then it looks like looking for a better machine is a must do first step or all else could be a waste of time.

A Line TBC cleans up the image (straightens out the frame and helps fix the colour errors). The trouble with buying another VCR is you don’t really know if the problem is the tape. If it is, you may well not get any improvement unless you fluke a high-end S-VHS machine that doesn’t have issues. A DVD recorder such as the ES-10 or ES-15 is relatively cheap and has no moving parts (the DVD part is not used) and will in most cases help a lot in fixing these types of problems. Whether it would actually fix that herringbone pattern, I’m not sure.

Do you think RYVT will have enough tools to help clean up the files further if needed after that purchase?

If it’s anything like MMS, I suspect not. While it looks like it can adjust colours (JohnEB’s comment) it probably has no noise reduction capability, like MEP doesn’t. That’s where Neat video is a godsend in this type of scenario.

I don't know if RYVT has the same ability to produce as high a quality MXV file as MEP.

Neither do I. As I suggested, the capture MXV quality should be as high as possible.

@cmclernon uploaded the MP4 file version which they have stated was the poorer quality of the export formats.

Weirdly, RYVT specs show that it can’t export H264 but it can export MPEG-4. For a video company, that’s pretty shabby documentation. Not being able to export H264 would be a deal-breaker for me.

Would it be possible to view the MXV file in MEP to see if the file size would be comparable to the MXV file sizes MEP can produce?

Would there be any advantage of recording the file to AVI? Is it possible in RYVT?

It can’t be done in MMS. I do not use MMS to capture these days, although early captures (2010ish) into MXV were fine. MXV at User Defined 100% quality is what you’d aim for.

I noticed where there is a relatively clean frame in the sample file that the MP4 has blocked whole areas of colour. Possibly due to the lack of a colour pallet from the VHS medium.

That is what is concerning me. Where there is not much detail or lack of interference patterns, the colour areas are being rendered in (relatively) large blocks. I think some sharpening or contrast adjustment at that stage may help if the mxv file has enough resolution or am I wrong? Also the exported final files needs to have less compression applied to my mind.

I think that would be a result of the capture quality and/or the export bitrate. I’m not sure sharpening or colour (brightness/contrast) adjustment would clear those blotches. And it could be the tape.

Would that be possible with RYVT?

Has MEP the ability to capture an analog video stream any better? Would it be worth the upgrade?

At this point in time, not from a purely capture point of view (not knowing what options are available for the MXV quality). But MMS looks like it has much more capability when editing (has a proper timeline) and the ability to apply Neat Video noise reduction (costs) and also H264 exports. BUT, the Section function is bugged.

I don't know yet whether cmcleron's goal is to put these files onto DVD / Blu-ray or to keep them as Digital computer files for better quality playback.

I would really like to see the mxv file uploaded to see if it is clean enough.

I know that RYVT has a much lower machine specification to run than MEP and that we have not asked cmcleron for their machine specification yet or whether they want to pursue this option.

Capturing is not actually that intensive (as long as you are not capturing direct to MPEG 2). Because it’s only 720x576/480, that is why the suggested machine specs are lower.

It could well be that if the file has enough data in it that recording the same off of a better VCR and then exporting the file at a much higher bit rate may be enough without having to go through more steps to clean up the image if some of my assumptions are correct above and Al verifies my conclusions.

Depends on how many tapes and how much money you've got. Cost-Benefit-Analysis and all that! Certainly, a S-VHS VCR should give a better picture, especially if it has an inbuilt TBC, but you have to spend money. I don't know where @cmclernon resides, but vcrshop has been recommended. As I say, my first step would be to get an ES-10, or 15.

Unfortunately, the quality of analogue video is directly proportional to the money thrown at it, and the results are all in the eye of the beholder. I'm happy with my cheapish VCR + ES15, with Neat Video applied. Others would treat my efforts with utter contempt! In this binary world, analogue video is very much a subjective thing.

First, get that MXV quality up to 100% and do an MPEG-4 export at high quality (bitrate of 6,000kbps) to see if that looks satisfactory.

@AAProds

Hi Alwyn

. . . . RYVT specs show that it can’t export H264 but it can export MPEG-4. . . . . Not being able to export H264 would be a deal-breaker for me. . . .

MPEG-4 is h.264/AVC - see here.

I will have a look at my Panasonic DVD/Tape VHS recorder to see how good it is transferring tape to DVD direct.

John EB

@johnebaker

John, I assume you checked the specs for RYVT? If not, here is what Magix says about it's own RYVT:

@AAProds

Hi Alwyn

I can see what you are saying, however if the 'MPEG-4' codec is not AVC/h.264 then what is it?

In an older version of RYVT specs it was referred to as MPEG-41 - is this a typo and meant to be MPEG-4 10 aka AVC?

This is where using a container file format to refer to a specific codec has always been an issue, and in the early years of h.264, AVC was often considered as a separate codec even though it is a 'subset' of h.264.

John EB

@AAProds

Hi Alwyn

I have raised the RYVT specs issue with Magix for clarification.

John EB

@AAProds @johnebaker @cmclernon

Hi Guys.

I'm sorry to be such a pain but I now have a few additional observations and questions.

They have come about because of the second topic also covering this issue where Al thankfully correct one of me mistakes regarding which field of interlaced footage comes first.

The supplied video has bottom field first selected, not top.

Maximum bit rate                         : 8 000 kb/s
Width                                    : 720 pixels
Height                                   : 576 pixels
Display aspect ratio                     : 4:3
Frame rate                               : 25.000 FPS
Standard                                 : PAL
Color space                              : YUV
Chroma subsampling                       : 4:2:0
Bit depth                                : 8 bits
Scan type                                : Interlaced
Scan order                               : Bottom Field First
 

When I look at the video very slowly frame by frame in MEP it surprised me I could see the two interlaced fields shown separately before the scroll wheel recorded the movement of one frame.

In every instance there was more of the cross hatching in the second field than the first field and in most instances where there was cross hatching in the first field it was often in different sections of the frame to the first field.

Would swapping to top field first during capturing the file cure some of that?

If the file was captured correctly but exported incorrectly, would that have caused the cross hatching to appear?

If anyone is capturing multiple video clips for use in the same project should field selection remain a constant regardless of source?

What happens if mixed processed clips were to be exported from the same project?

What is causing that cross hatching is bugging the heck out of me. I'm wondering if it actually shows up in the mxv file or has been amplified by the compression of rendering to MP4.

I tried cleaning the MP4 file with a trial version of Topaz Labs AI Video enhancer and the main thing it could not cope with was that Cross hatching.

I can put up copy of that conversion if later when @cmclernon has seen these recent posts, they have no objection to the clip being uploaded to this topic.

Ray.

@CubeAce

Ray, I doubt the crosshatching is the result of mixed-up field order. I deinterlaced the MPEG file in Virtual Dub and the crosshatching is still there, in every frame.

For what it's worth, MPEG 2 is normally BFF. I don't know why Magix persists with providing the field order option for export. Interlaced MP4 is not the norm either.

I don't think mixing field order is an issue in a project.

@AAProds

Hi Al.

I think there may be a misunderstanding here.

Surely once an artifact is encoded into a file it can't then be filtered out in the ordinary sense.

This is one reason I would like to see a copy of the mxv file rather than an MP4.

I am at a loss to explain why if the video tape is shown on a TV it is not showing up (or hardly noticeable) but showing up on an exported digital file.

The only other things I can think of are.

A poor signal path caused by either overly long leads or bad connections (oxidisation of lead or socket contacts or poor shielding.)

Or.

A localised strong magnetic field possibly caused by leads or machinery being in close approximation to something like loudspeakers or headphones placed nearby.

Both explanations I'm not overly keen on as alternatives as I think that would have hit the audio as well.

I've tried internet searches on analog transfer problems adding visual cross hatching and other search parameters so far without success.

Even the research papers such as this one I've found for digital archiving of VHS tapes have not mentioned that as a problem. I thought that one was a good paper though.

Still, thanks for your valued thoughts and experience.

Is it possible that because in theory, a file could be used for upload to a file sharing site at a higher quality than possible for normal DVD / Blu-ray playback that is why Magix has included export to MP4?

Ray.

@AAProds

Hi Alwyn

. . . . I don't know why Magix persists with providing the field order option for export.. . . .

The most likely answer is because not all MPEG 2 are BFF, My first Sony camera was TFF - recorded direct to hard drive.

. . . .  crosshatching is the result of mixed-up field order . . . .

RFI when capturing? The crosshatching it is not consistent to indicate an issue in the VCR - which could be the Daewoo Q812Pi - a 6 head VCR.

John EB

@CubeAce

Is it possible that because in theory, a file could be used for upload to a file sharing site at a higher quality than possible for normal DVD / Blu-ray playback that is why Magix has included export to MP4?

I don't think so. The only reason Interlacing was invented was to reduce the data rate for the early TV systems. It wasn't an issue of size, per se. It is universally reviled by everybody these days as being a total PITA (IVTC, 3:2 Pulldown, it's a nightmare to understand and do something about, fortunately PAL is much simpler). Online streamers such as Youtube send you Progressive video.

That paper has a good section on TBCs, but otherwise is well out of date. "Proper" analogue capture is done into a lossless codec such as HuffYUV, Lagarith or even Magix MXV (at maximum quality settings). That keeps the very best quality for the restoration process and eventual export out to MP4 or for further editing or archiving. DV capture, especially NTSC, is frowned-upon by the aficionados, as is MPEG 2 for capture. Both are already lossy formats and while you certainly can use them for capture ( I have done it in the past) it's not the "enthusiast's" way of doing things.

@johnebaker

The most likely answer is because not all MPEG 2 are BFF, My first Sony camera was TFF - recorded direct to hard drive.

John, I think it's more that Magix is not keeping it's eye on the ball. It's as though they created the program years ago and now they have just given up (our comments re the specs is but one example). You won't find many applications for interlaced MPEG 2 or MPEG 4 these days (except DVD, which automatically creates the field order for you). For example, a lot of exporting in other programs is done using CRF, while we are still stuck with guessing the bitrate. The only thing that seems to have improved is the (previously slow) reading and encoding speed.

@AAProds

Hi Alwyn

Re RYVT / Video Easy:

. . . . 'I think it's more that Magix is not keeping it's eye on the ball.' . . . created the program years ago . . .

The question is how do you improve a relatively simple program that does what is does at the price point it is at and specifically targets obsolete media digitisation, with a decreasing market.

Interlaced versus progressive - there are Pros and Cons with both and that discussion will go on for years with no conclusion other than Interlaced will die out as the last analog viewing devices are gone and all broadcast is progressive.

. . . . exporting in other programs is done using CRF, . . . .

Any program that exports h.264/AVC and h.265 is using CRF parameters, they are an inherent requirement for the codecs. The CRF values are optimised in the export preset settings for best perceived visual quality vs bitrate/file size and should not be touched unless you have the analysis tools to determine what is happening 'under the hood'.

. . . . . while we are still stuck with guessing the bitrate . . . .

IMHO we are in a better situation as we can adjust the target (average) and allowed maximum.

Using CRF as the controlling factor in export is guessing, if you want a specific bitrate and are trying to control it with adjusting CRF you do not know what the final bitrate will be, ie you would have export, determine the bitrate, adjust the CRF export again and rinse and repeat until the desired bitrate is achieved.

Depending on the video content, 2 very different content videos will have different bitrates for the same CRF, however the perceived quality will be the same.

John EB

I seem to have started quite a discussion here! I'm not really a video techie myself so first of all what is MEP, CRF and MMS?

My main objective is to remove the cross hatching flickering/banding patterns so as to produce an exported result I can be happy with. Would the Panasonic ES15 as a passthrough device achieve this. Would all DVD / HDD recorders have TBC in passtrough? (I've seen some very reasonably priced HDD recorders with Freeview on EBay which could therefore still be useful apart from being TBCs)

I no longer have the MXV file for the sample I've uploaded but please find a link to the MXV file of another test sample which shows that the bands/crosshatching are already there at the capture stage:

https://drive.google.com/file/d/1EuOaOdv29IpSe0ESPlRrDKn--nxvvFpF/view?usp=sharing

I export as an MPEG2 file rather than an MPEG4 file partly as I've said, because the letter's quality is much poorer than the former, but also because I can't deinterlace the latter to produce smooth motion video.

On a separate topic, my exported files are rather blocky, though they look satisfactory on my TV. Nevertheless, I would have thought that over 4 Gb for a 90 minute video was large enough.

@cmclernon

Hi.

Sorry but that is another MP4 file not an mxv file. MXV files will only work in a Magix editor and not a media player or other video editor. mxv files will not show up with thumbnails within Windows.

Looking at the file within a Magix editor I can see the file size and compare it to other file formats or the same format but generated using Movie Edit Pro of Video Pro X.

Also, unlike an MP4 file or similar, there should be no blocking of colour areas.

I think the jury is out as to whether a TBC is going to cure the cross hatching. It is embedded in the file even in areas not showing in normal viewing.

The full sized file can be downloaded from here.

I took the chroma and luminescence completely out of the video and this is what is left behind. These are extracted from Bitmap still images of one of the frames from the MP4.

I think it looks to me (my own personal opinion not based on science but more of a gut feeling based on similar problems when dealing with noise with older bridge camera stills raw files) like it could either be sensor noise due to filming in artificial lighting conditions as there doesn't seem to be any additional lighting except the room lighting or it could be the capture file if you used xmv for capturing the files and yet it is still not capturing enough data.

@AAProds

Hi Al.

Just a thought on the other problem.

If you have New Blue Essentials, there is a Crop Boarder effect that holds on multiple instances of use.

Ray.

@cmclernon

Regarding files sizes for digital capture of the VHS analog signal.

To my mind I would want to use a non lossy codec that would give each individual pixel when digitized it's own set of values and not be grouped with its nearest neighbors value and become a block of colour, no matter how small that block would be. I don't know what RYVT options there are for screen capture from an analog source. I thought that mxv is supposed to be a Magix version of a non lossy format. If that is true then why is there a quality slider for it within Movie Edit Pro?

That may be a rhetorical question, I'm not sure. I don't even know what Movie Edit Pro offers in the way of options for analog screen capture as I have no analog recordings to try out.

If on the other hand you are recording in RYVT directly to MP4 or mpeg-2 before editing, that would be a mistake.

Ray.

@CubeAce I've just noticed that the video capture file used is indeed MP4 when I'm capturing from VHS using a composite input. When I'm capturing from a Hi8 tape using an S-Video input, the capture file is MXV and there is no crosshatching in the export file. I wonder could this be the answer,

@cmclernon

That seems a plausible theory to me if there are no options to choose from although it could be the recording to MP4 is just making the cross hatching more visible due to the compression used in MP4 rather than the cause of the cross hatching.

Lets see what John and Al make of it. They at least have had dealings with recording analog files.

@AAProds @johnebaker

Just adding them to make sure they get notification of this additional information.

Ray.

No I've just found that most of my Hi8 capture files are MPG (not MP4) as well. Don't know where the MXV file came from.

I've just noticed that there is no blockiness in the export file created from the MXV capture file, unlike those created from MPG capture files.

How do ensure that the capture file is MXV?

@cmclernon

I'm back. 😉 Could you download the Video Easy PDF manual (for Movie Studio, it's on the Help menu>Download PDF manual) and post it on Google Drive for us.

I'm assuming that the actual program you're using is Video Easy; the name of the package itself is Rescue Your Video Tapes because it includes the capture stick.

The workflow for analogue captures with Movie Studio (Video Easy's big brother) is capture in MXV, then do your editing, then export as MPEG-2 or 4, and this is what I'd recommend for Video Easy, but we'd need to see the manual to see what options you have.

Re the cross-hatching, it is certainly feasible that when capturing Hi8 via S-Video, you won't get it. The VHS captures have the tape and the VCR as variables.

what is MEP, CRF and MMS?

Sorry about those! MEP=Movie Edit Pro, CRF=Constant Rate Factor (tangent, ignore) and MMS=Magix Movie Studio. Movie Edit Pro was the name for decades of Magix's middle of the road video editor. For one reason or another, Magix changed it's name recently to Magix Movie Studio, so you'll see us reverting back to the old name at times.

My main objective is to remove the cross hatching flickering/banding patterns so as to produce an exported result I can be happy with.

I can't be sure about removal of crosshatching, but what I am sure about is that a TBC (DVD Recorder in passthrough) will fix the wobbly top section and the iffy colours. If you step through the bit where the gentleman is sitting in the chair before the lady comes in, the picture frame, curtains and door frames really are very wobbly.

Would the Panasonic ES15 as a passthrough device achieve this.

Yes, the ES-10 and 15 are the gold standard here.

Would all DVD / HDD recorders have TBC in passtrough? (I've seen some very reasonably priced HDD recorders with Freeview on EBay which could therefore still be useful apart from being TBCs)

No. Only some machines do this, mainly early-model Panasonics. Ones I know that do do it are the DMR-ES10 and ES15, the DMR-E55, DMR-E60. Also Pioneer DVR-440, Sony HXD 870 and 890.

The Panny DMR ES35 and DMR EZ48 are VCR/DVD combo machines that have TBC built-in, with S-Video Out. It is as good as the ES10 and 15.

Bear in mind that for the DVD recorders, you don't need the DVD function to work; you're only using the passthrough feature.

One caveat: all my experience is with PAL gear so I can't vouch for the above in NTSC land.

If you have another model in mind, let us know what it is and we can research it for you.

I would have thought that over 4 Gb for a 90 minute video was large enough.

If it's MPEG-2, that would only just be enough. Given the quality of your tape, I'd be inclined to crank up the bit rate even more, if you can. Also, export as MPEG-4: less size for the same quality.

I export as an MPEG2 file rather than an MPEG4 file partly as I've said, because the letter's quality is much poorer than the former, but also because I can't deinterlace the latter to produce smooth motion video.

The quality shouldn't be less unless the export bitrate is lower.

In the second sample (Recording - 0001.mpg) I think the camerawork is the main reason for the jittery video. And again, the wobbly top half, and flashing colours would be corrected by a TBC IMO.

The crosshatching is still there on the dinner table MPG; is that the same tape as the man in the chair? If so, try capturing a different tape. That might isolate the crosshatching.

If you can get your hands on another VCR for a quick check, that would further isolate the issue.

If you're brave, you could rip the top off your VCR and clean the heads, but only with printer paper soaked in isopropyl alcohol (IPA). Do NOT use cotton tips under any circumstances. Let me know if you want to try that. It's not particularly difficult.

@AAProds @cmclernon

Just a quick word on head cleaning.

Do not use any isopropyl alcohol as it is not pure enough. Normally isopropyl alcohol pure enough for tape head cleaning of film cleaning is at least 99.9% pure and you get it in 250ml and upwards bottles or cans.

There are a few synthetic sponge and felt swabs that could be used as well as micro-fiber lint free cloths.

If you are editing video, as in cleaning up images or cross fading clips rather than hard cuts, the capture files should be as high quality as possible or else you will be compressing the video twice when exported.

All I can find on RYVT is this.

Ray.