VideoHelp Forum
+ Reply to Thread
Page 2 of 2
FirstFirst 1 2
Results 31 to 51 of 51
Thread
  1. Yes, I don't know what video there is here.
    First, I understand that you are something like half God of making video pretty, nevertheless,

    I'd say NeatVideo delivers and is was designed for those with no knowledge about denoising and denoise, choosing area, profile is created and done. This is good for internet, where you usually resize also, it brings down video to half size for uploads, if it is your server not YouTube video quality is MUCH better because there is bandwith limit. YouTube tends that difference bring down re-encoding again. Sure as a back up someone chooses CRF 16-18 and it is encoded, as is, without denoise. BUT for low-light shots after denoise with Neat video it looks better also, even using Sony VX2000, which was considered as a queen among consumer DV avi camcorders, in low light it helps. Originals are still kept as is without denoise, so no problem. For example smart render DV avi is done very fast, literally seconds to get final little movie as is, and then denoised, processed to get mp4 out of the same NLE timeline, you can even save two projects before and after processing. Nowadays HD camcorder video downsizing for web gives better overall result, where usually no deinterlace is needed as well.

    If somebody makes video very often (every week) needs something fast if needed to fix noise, there is no time to spend hours to tune up denoise filters. You cannot processes clip separately outside of NLE and then bring it there again. You do that and you have divorce on your hand because all weekends are wasted on some video

    I forgot main point,..., if you go lossless than you HAVE TO encode, if you keep DV avi in NLE you can smart render your movie.
    Last edited by _Al_; 1st Sep 2013 at 12:15.
    Quote Quote  
  2. Banned
    Join Date
    Oct 2004
    Location
    New York, US
    Search Comp PM
    Originally Posted by _Al_ View Post
    Yes, I don't know what video there is here.
    First, I understand that you are something like half God of making video pretty, nevertheless,
    Not true. I just follow establish practice. No one here is a god of anything.


    Further modifications for uTube will obviously altwer the video in a number of ways, and uTube will re-work everything anyway to suit their purpose.

    I use NeatVideo all the time -- if, and only if, the source requires it. DV source is usually fairly clean and wouldn't need something like NeatVideo unless it's been damaged or made noisy somehow. But it depends on the source.

    You can't color correct DV without decoding and then re-encoding it.
    You can't denoise DV without decoding and then re-encoding it.
    You can't deinterlace DV without decoding and then re-encoding it.
    You can't change DV colorspace without re-encoding it.
    You can't resize DV without re-encoding it, and you should never resize interlaced video.
    Each time these separate operations are performed, the video will be decoded and re-encoded.

    Considering all that the owner proposed to do with the videos, I'd recommend decoding only once, performing the operations all at one time, and re-encoding after all intermediate processing is completed. If the O.P. wants to feed it all through Vegas and perform all of that at one time, it's up to the owner.
    Last edited by sanlyn; 21st Mar 2014 at 15:37.
    Quote Quote  
  3. Touching DV avi in any way gets compressed video re-encoded that is clear, this is true for any compressed video (not lossless), sure, but having that DV avi on timeline you can easily get your movie out after editing as DV avi as well, original, then you add titles, logos, color correct to get , deinterlace if you want to get mp4. The most work while editing is making movie, shuffling clips around, syncing it, getting all sound as -6dB or so , sound transitions are more time consuming to produce than video transitions. Adding titles and possible denoise or color correct is just final procedure. I was (am) practicing this all the time.

    If he is using external programs, you should go lossless, definitively, but you see after smart render. That DV avi could be used in that middle state after smart render as well as for whole movie or just parts of it (couple of parts, not tens of clips, that's time consuming), he can load that smart DV avi render into After Effects or other applications exporting lossless in that state.
    Quote Quote  
  4. Originally Posted by _Al_ View Post
    oh, it is DV video, which is standard definition, so there should be:
    Code:
    ConvertToYV12(interlaced=true, matrix="PC.601")
    matrix="PC.601" turns light blue sky into white; the same happens if I open the source clip in mpc-hc

    after converting back to yv12 using matrix="Rec601", it looks like it looks in vegas preview (sky is light blue, not white)

    Click image for larger version

Name:	aver0000.jpg
Views:	376
Size:	70.3 KB
ID:	19860

    Above is a source clip opened in avisynth. In the top right of the avisynth histogram, is the sky. It looks like it exceeds 235, which can be fixed with levels filter in vegas, but if I understood sanlyn correctly, it is better to fix it in yuv with avisynth, before it gets converted to rgb after it is imported in vegas or premiere

    Click image for larger version

Name:	levels.jpg
Views:	248
Size:	228.4 KB
ID:	19861
    Last edited by codemaster; 1st Sep 2013 at 16:04.
    Quote Quote  
  5. Why did you not made it clear that you are not using DV avi? I even asked question about it, you continued in discussion as if confirming that. Men, I hate this . I never edit mpeg2, never recorded anything with mpeg2 camcorder, would never know about camcorder coloring and Vegas's behaviour with that format.

    Post a short clip. The whole clip, not cut off clip somewhere, let see if Vegas can smart render those.
    Last edited by _Al_; 1st Sep 2013 at 14:58.
    Quote Quote  
  6. Ok, I've uploaded a clip. Mpeg2 is also used as a capture codec. It is a valid capture codec, just like uncompressed, dv, motion jpeg, dnxhd, prores, cineform, huffyuv, lagarith, ffv1, avc-intra. Vegas and Premiere have native support for mpeg2.

    I'm trying to figure out what sanlyn explained so far. I'll try to process the source clips in yuv (before the rgb conversion), like he recommended, then compress to lossless lagarith.
    Image Attached Files
    Last edited by codemaster; 2nd Sep 2013 at 01:15.
    Quote Quote  
  7. Banned
    Join Date
    Oct 2004
    Location
    New York, US
    Search Comp PM
    Originally Posted by codemaster View Post
    Ok, I've uploaded a clip. Mpeg2 is also used as a capture coded. It is a valid capture coded, just like uncompressed, dv, motion jpeg, dnxhd, prores, cineform, huffyuv, lagarith, ffv1, avc-intra. Vegas and Premiere have native support for mpeg2.

    I'm trying to figure out what sanlyn explained. I'll try to process the source clips in yuv, like he recommended, then transcode to lossless lagarith.
    We need more detgail about your source. If your camera records MPEG2, then likely it is already YV12. What camera? What format odes it record?

    MPEG shouldn't be captured as another encoded MPEG. You should be able to copy MPEG to your PC. If you brought it into your PV as DV, you should know that MPEG and DV-AVI are not the same thing. Capturing MPEG to DV-AVI is a re-encode.
    Last edited by sanlyn; 21st Mar 2014 at 15:38.
    Quote Quote  
  8. Camera is Canon DC100. It writes vob, bup and ifo files on optical disc (miniDVD). Camera writes all clips into one or two vobs, all clips into one continuous VTS, separated with chapters. Vegas Pro is able to ingest the clips directly from this optical disc (File->Import->DVD Camcorder Disc...), is able to figure out where a clip starts and ends, copies it's original streams and wraps them in mpg container. No transcoding is necessary. Any other software that can do this, is not able to maintain the audio properly synchronized.

    The simplest way to edit is to use Vegas native support for this camcorder. But now I understand that it's better to deinterlace and fix levels first (before the yuv->rgb conversion), then compress to lossless, and then edit.
    Last edited by codemaster; 2nd Sep 2013 at 01:27.
    Quote Quote  
  9. Banned
    Join Date
    Oct 2004
    Location
    New York, US
    Search Comp PM
    You're correct so far, especially since your vob sample is rather noisy. You can copy a VOB directly to your hard drive. You can use VOB2MPG to copy multiple VOB's off disc and join them as a single MPEG, there should be no problem with audio sync. It has some edge noise, motion noise, and aliasing. Nice, accurate color. Levels look good as well. A bit soft (it's a 2-hour bitrate). Thanks for the info and the sample. Readers can have a look at something more specific now, and offer more details suggestions.

    I'll be able to give your sample a more thorough look this evening. So far, all I've done is to make a d2v project using DGIndex, and demuxed audio into uncompressed PCM.
    Last edited by sanlyn; 21st Mar 2014 at 15:38.
    Quote Quote  
  10. Vegas 12 seemed to recognized upper field first. But good news. Vegas CAN smart render this video. Meaning you can load it, edit render out without change. Only if video will not start with I-frame or you brake GOP at any point - crossfades, cut the end, beginning of any clip, those parts will be rendered again.

    So you can edit your video, then smart render - MainConcept mpeg2/Program stream PAL widescreen , then you change field order to upper field and set CBR as 9,400,000 (I took this value from Media Info - maximum bitrate and Vegas seems to render smart then). Then you can fix the whole movie, colors, noise etc outside of Vegas, still YUV, same video and as a lossless you decide what to do with it.
    - Importing back to Vegas , finishing touches titles, but you know you'd have to go YUV-RGB-YUV route
    - you just encode it. Without any YUV-RGB-YUV processing (except those parts without smart render). It's up to you what you decide. If yo have very long clips in your video it would not matter at all.
    - You can make DVD right away this way without re-enconding it at all if you do not fix your video after smart render.

    I'm sure guys here can fix your video with their magic, it is open ...
    As I said I would not go lossless since beginning, as Richard Pryor would say, "Your choice, ... your video, your choice", I know I still mess up Sanlyn's workflow, sorry for that.

    About that dmfs, I tried dmfs out of Vegas and got the same colors using:
    Code:
    AviSource("input.avi")
    ConvertToYV12(interlaced=true, matrix="PC.601")
    Spline36Resize(720,404)
    just test it, but the script makes no sense, resizing interlace, I guess you intended to go like this:
    Code:
    AviSource("input.avi")
    ConvertToYV12(interlaced=true, matrix="PC.601")
    assumetff()
    QTGMC() #you can denoise here too
    selecteven()
    Spline36Resize(720,404)
    Last edited by _Al_; 1st Sep 2013 at 21:18.
    Quote Quote  
  11. Originally Posted by codemaster View Post
    matrix="PC.601" turns light blue sky into white; the same happens if I open the source clip in mpc-hc
    DV camcorders very often record out of spec. Often exceeding Y=235 at the top end, and have black levels too high at the bottom end. So you have to be careful to adjust levels before converting to RGB. And be sure to use a DV decoder that can output YUV so the decoder itself messes up the video (Panasonic DV Codec only outputs RGB and crushes brights like in your sample). I recommend Cedocida.
    Quote Quote  
  12. Originally Posted by jagabo View Post
    DV camcorders very often record out of spec. Often exceeding Y=235 at the top end, and have black levels too high at the bottom end. So you have to be careful to adjust levels before converting to RGB.
    I think sanlyn said something similar. He recommended to first fix the levels, so that they don't exceed 16-235, compress to lossless yv12, and then import into NLE, so that the conversion to RGB will not clip levels when it maps 16-235 to 0-255. Or convert to rgb with avisynth, and then import into NLE, perhaps because the yv12->rgb colorspace conversion is of better quality in avisynth than in NLEs.
    Last edited by codemaster; 2nd Sep 2013 at 01:32.
    Quote Quote  
  13. I've got a few questions about sanlyn's procedure at post #28.

    First of all, my target is web or PC display always, and never DVD.

    At step 2, you mention opening source in avisynth and decoding it to avi lossless yv12, but does this include (in that avs script) deinterlacing, denoise and fixing levels if yuv range exceeds recommended values ?

    At step 3, those preliminary color corrections are made with avisynth histograms, vectorscopes, and Levels() function? After deinterlacing, or before? Also, what is the proper processing order: first fix levels, then fix colors, then deinterlace, then denoise? Or: deinterlace -> denoise -> fix levels -> fix colors ?

    At step 4 and 5, you are suggesting that, after step 2 and 3, I open the YV12 avi in avisynth, convert it to RGB, save it to a new avi compressed with lagarith with mode RGB, and then open this RGB avi in the NLE? And then, after editing, color correcting, adding fades, titles, etc, to render to avi RGB uncompressed or losslessly compressed, and convert to yv12 using avisynth? This means YV12->RGB->YV12 with avisynth instead of with NLE? Why? Less information lost through the colorspace conversion? You also mention there are sophisticated avisynth plugins for yv12->rgb conversion. What plugins are these, what are their names? You meant sophisticated plugins just for yv12->rgb, or for yv12->rgb and rgb->yv12?

    At step 6, apple aac is higher quality than nero aac, at vbr 160kbps? x264 is good enough, or there is a higher quality h.264 encoder?
    Last edited by codemaster; 2nd Sep 2013 at 00:48.
    Quote Quote  
  14. Banned
    Join Date
    Oct 2004
    Location
    New York, US
    Search Comp PM
    https://forum.videohelp.com/threads/309449-Adobe-Premiere-MainConcept-s-H-264-vs-x264-o...source-encoder
    MainConcept vs x264 encoders

    comparusons x64 -vs- others
    https://forum.videohelp.com/threads/308482-high-quality-mpeg-2-encoding/page5


    Originally Posted by codemaster View Post
    I've got a few questions about sanlyn's procedure at post #28.

    First of all, my target is web or PC display always, and never DVD.

    At step 2, you mention opening source in avisynth and decoding it to avi lossless yv12, but does this include (in that avs script) deinterlacing, denoise and fixing levels if yuv range exceeds recommended values ?
    Ye.In the case of your sample VTS01q mpg, the levsl and chroma are OK for both PC and NTSC/PAL standard.

    Originally Posted by codemaster View Post
    At step 3, those preliminary color corrections are made with avisynth histograms, vectorscopes, and Levels() function? After deinterlacing, or before? Also, what is the propeF processing order: first fix levels, then fix colors, then deinterlace, then denoise? Or: deinterlace -> denoise -> fix levels -> fix colors ?
    fix the luma and color range, if necessary. Then deinterlace, then denoise. Some scenes might require level and color tweaks; this is especially true of VHS, which has such screwey color rendition that RGB and complex filters like gradation curves might be needed. But if all is well up to that point, there's no need to get complicated. And note that this sequence is sometimes altered for really grungy dirty damaged horrible video. Fortunately, that's not quite the case here.

    Originally Posted by codemaster View Post
    At step 4 and 5, you are suggesting that, after step 2 and 3, I open the YV12 avi in avisynth, convert it to RGB, save it to a new avi compressed with lagarith with mode RGB, and then open this RGB avi in the NLE? And then, after editing, color correcting, adding fades, titles, etc, to render to avi RGB uncompressed or losslessly compressed, and convert to yv12 using avisynth? This means YV12->RGB->YV12 with avisynth instead of with NLE? Why? Less information lost through the colorspace conversion? You also mention there are sophisticated avisynth plugins for yv12->rgb conversion. What plugins are these, what are their names? You meant sophisticated plugins just for yv12->rgb, or for yv12->rgb and rgb->yv12?
    Basically, that's correct. Colorspace conversion depends on the video, but I maintain that AVisynth can do a better job of it. "Special" plugins for this work might be required -- again, depending on the video. Nowadays I usually use plugins such as SmoothAdjust and the dither() package, which I think keeps colors looking clean and definitely helps minimize data loss. They work within (or simulate) a 16-bit space instead of the usual 8-bit. But that's up to you. Most of the time you can make do with AVisynth's methods. Again, it depends on the video. Those special filters vare also used to help prevent banding and other artifacts.

    Originally Posted by codemaster View Post
    At step 6, apple aac is higher quality than nero aac, at vbr 160kbps? x264 is good enough, or there is a higher quality h.264 encoder?
    I won't be doing any of this with lossy aac or mp2 audio. I decoded your mpg to lossless YV12 and decompressed the audio to PCM (.wav). Re-encoding the audio would occur during final video encoding. X264 is certainly among the best of encoders. Vegas and Premiere don't use x264. VBR for audio almost always causes problems.

    As for the VTS01 mpg sample: it's off to a poor start. Correct field order aside, I wonder how your sample was made? If it was cut and saved in Vegas, it was re-encoded. There are several problems: aliasing, dot crawl, broken or ragged lines and edges, edge ringing, and some rather weird problem that looks like field displacement. I invite some gurus in the interlaced field placement area (who know much more than I do about it) to have a look at the original MPG and the demo below.

    I'll have more detailed images later, but note the image below. The top picture is a reduced-size frame 450 from the interlaced original. Even with that small image, you can see dot crawl and sawtooth edges in the bright part of the small pointy cone-shaped object in the lower right-center. This is not interlace combing. Later I'll post bigger samples. For now, a yellow arrow points to the small area that I have enlarged 3X in the two smaller pictures. After deinterlacing, the lower and upper borders flicker and change shape at alternating frames. The Even frames have a slight amount of discoloration along the bottom. Odd frames have a black border and about 8 vertical pixels of geometric distortion (ripple).
    Image
    [Attachment 19867 - Click to enlarge]


    The attached mkv has 3 deinterlaced versions of the mpg. Deinterlacers were the only filter used. The 3 15-second versions play consecutively. Each version is captioned in the upper left corner of the video. Version 1 (yadif), version 2 (QTGMC at "very fast"), version 3 QTGMC at "medium"). As the video plays, keep your eye on the top and bottom border flutter for all 3 versions. If anyone can explain this border problem, I'd appreciate it.

    I used three simple scripts for each version, with the upper left caption added later. All were saved as Lagarith YV12, 720x576, encoded for 16:9 display with an x264 encoder. No other filters were used. At this point, no audio.

    Code:
    MPEG2Source(videopath+"VTS01.d2v")
    yadif(Mode=1,order=1)
    
    MPEG2Source(videopath+"VTS01.d2v")
    AssumeTFF().QTGMC(preset="very fast",sharpness=0.8)
    
    MPEG2Source(videopath+"VTS01.d2v")
    AssumeTFF().QTGMC(preset="medium",sharpness=0.8)
    Last edited by sanlyn; 21st Mar 2014 at 15:39. Reason: I can't possibly be making all these typos. Gremlins are doing it!
    Quote Quote  
  15. sanlyn, did you recommended to:

    a) open source in avisynth -> compress to lossless yv12 -> open the compressed lossless in avisynth -> fix levels if necessary, deinterlace -> compress to lossless yv12

    or

    b) open source in avisynth -> fix levels if necessary, deinterlace -> compress to lossless yv12

    Is it a or b? I am assuming a means "working in lossless YV12", and b means "working in lossy YV12". If what you said is b, then what is the reason for fixing levels and deinterlacing in lossless YV12 ? What is the difference or advantages of fixing levels and deinterlacing using lossless YV12 versus fixing levels and deinterlacing using lossy YV12 ?
    Last edited by codemaster; 2nd Sep 2013 at 17:04.
    Quote Quote  
  16. Banned
    Join Date
    Oct 2004
    Location
    New York, US
    Search Comp PM
    Do everything in the original colorspace unitl it's time to do something else that requires another colorspace for processing -- such as Vegas. In the short script I posted, I would not have had to use Vegas at all because DGindex works directly from VOB. I used the free DGindex utility to create an project file (or index) to an mpeg2-encoded source and to convert its AC3 audio to uncompressed PCM. The script didn't work with audio because it wasn't needed for the demo. I opened the script in VirtualDub, checked the reults, and saved the AVI with VirtualDub in this manner, without changing the original colorspace:

    In The VirtualDub top menu, click "Video..." -> "Color depth..." -> "4:2:0 planar YCbCr (YV12)" -> "OK".

    In The VirtualDub top menu, click "Video..." -> "Compression..." -> "Lagarith Lossless Codec"-> "Configure" -> "YV12" -> "OK" -> "OK"

    In The VirtualDub top menu, click "Video..." -> "Fast recompress".

    In The VirtualDub top menu, click "File..." -> "Save as AVI..." -> give the file a name and location.

    Voila! Decoded, deinterlaced, and saved as YV12 lossless compression with Lagarith.

    I could have added more to the script. Givea pause for a short dinner with the wife (women DO NOT understand the priority of these projects), and I'll demonstrate some of the things I tried with this sample video.
    Last edited by sanlyn; 21st Mar 2014 at 15:39.
    Quote Quote  
  17. Banned
    Join Date
    Oct 2004
    Location
    New York, US
    Search Comp PM
    Originally Posted by codemaster View Post
    sanlyn, did you recommended to:

    a) open source in avisynth -> compress to lossless yv12 -> open the compressed lossless in avisynth -> fix levels if necessary, deinterlace -> compress to lossless yv12

    or

    b) open source in avisynth -> fix levels if necessary, deinterlace -> compress to lossless yv12

    Is it a or b? I am assuming a means "working in lossless YV12", and b means "working in lossy YV12". If what you said is b, then what is the reason for fixing levels and deinterlacing in lossless YV12 ? What is the difference or advantages of fixing levels and deinterlacing using lossless YV12 versus fixing levels and deinterlacing using lossy YV12 ?
    Sorry. I neglected to reply in detail. When you open a file in Avisynth with AviSource, MPEG2Source, or whatever, the file is decoded while Avisynth handles it. The video remains lossless until it is submitted to another lossy encoder such as MPEG2, h264, DivX, XVid, DV-AVI, etc. Avisynth's output itself is lossless. Therefore, both a and b will process as lossless and output as lossless.
    Last edited by sanlyn; 21st Mar 2014 at 15:39.
    Quote Quote  
  18. Banned
    Join Date
    Oct 2004
    Location
    New York, US
    Search Comp PM
    The two images below were made from two frames in the VTS_01_1 (01).mpg sample from post #36. This mpg was the devil to work with. It should have been created with DGIndex from the original VOB rather than (apparently) cut/edit in Vegas and re-encoded. It has aliasing, dot crawl, brpoken and ragged lines, ringing, halos, moire/stairstepping and object shifting on motion, and other problems. It takes a lo of filtering to clean these disturbances, and one will never get rid of them entirely. Most of this is from lossy re-encoding in an NLE editor or could be faults of the camera itself.

    below: frame 485 of the original interlaced mpg. In the upper left is a reduced-size image of the frame. The other two images are 2X blowups of the areas indicated by the arrows.
    Image
    [Attachment 19886 - Click to enlarge]


    below: the same idea, from frame 915 of the original interlaced mpg. Even at a 2-hour bitrate of 4600 VBR, there should not be this many problems from a digital camera.
    Image
    [Attachment 19887 - Click to enlarge]


    In the second image you'll see how washed out the nackground looks. It also has a visibly different color balance from the start of the clip. It's caused by autogain and autocolor in the camera. The hills and other objects in the background are lighter than the same hills at the start of the video. It's a common problem with those auto features. There's little or nothing that can be done about it.

    The original mpg has a lot of edge noise every time something moves (including the camera) because dot crawl does exactly as its name implies: it "crawls" from frame to frame.
    Last edited by sanlyn; 21st Mar 2014 at 15:40.
    Quote Quote  
  19. Originally Posted by sanlyn View Post
    there should not be this many problems from a digital camera.
    Unfortunately, it's very common. Digital cameras usually don't capture the entire CCD image then scale it down with a nice resizing filter 30 times a second. They usually sub-sample the CCD -- essentially a point resize. That leads to exactly these type of aliasing and moire problems.
    Quote Quote  
  20. Banned
    Join Date
    Oct 2004
    Location
    New York, US
    Search Comp PM
    After a lot of experimentation and filtering (which should not have been required with a cleaner source), the two mkv's posted below are about as clean as I could get the original sample. This kind of cleanup reduces fine detail. I used low-power and medium-power filters to get there. I could be incorrect about the reason for all the artifacts in the mpg, but vids I've seen from similar Canon and Panasonic cameras don't usually have this many problems.

    A_VTS01_50p.mkv is 50FPS progressive PAL, 720x576 original size.
    B_VTS01_50p_720x404.mkv is 50FPS APLO, resixzed with Spline36Resize in lossless format to 720x404. The size of the two mkv's is approximately the same. Dome of noise cleanup returns as a result of resizing.

    The attached .txt (VTS01_script.txt)file is distilled copy of the script I used, with some comments. This script does a lot of work and runs at about 6 fps in VirtualDub, so it took about 6 minutes to clean this clip. A single avs script does all the work in lossless AVI, then converts to RGB for work in VirtualDub and the NLE editor. I added 3 transitions in lossless mode: a fade-in, a "roll" at about 20 seconds in, and a fade-out to black.

    This was the workflow:
    (a) Using one avs script, decode the d2v project file with DGIndex into lossless YV12 AVI, join PCM uncompressed audio to the video, adjust levels and bring up shadow detail a little, deinterlace, filter, then save as lossless RGB32 with Lagarith.
    (b) Open the lossless RGB clip in VirtualDub for light-power NeatVideo and a fade-in/fade-out, save as lossless RGB with Lagarith.
    (c) Open the lossless RGB in an NLE, cut the video in half and add a transition effect at the join, then encode with the NLE's x264 coder.
    (d) Open the same lossless RGB from step (b) with Avisynth and resize to 720x404, then open the lossless result in the NLE and do the same thing I did in (c).

    I used one script, no smart rendering, and one encoding step. All procedures were with lossless media.
    Image Attached Files
    Last edited by sanlyn; 21st Mar 2014 at 15:40.
    Quote Quote  
  21. Banned
    Join Date
    Oct 2004
    Location
    New York, US
    Search Comp PM
    Originally Posted by jagabo View Post
    Originally Posted by sanlyn View Post
    there should not be this many problems from a digital camera.
    Unfortunately, it's very common. Digital cameras usually don't capture the entire CCD image then scale it down with a nice resizing filter 30 times a second. They usually sub-sample the CCD -- essentially a point resize. That leads to exactly these type of aliasing and moire problems.
    Perhaps the few such cameras I've seen (3, at last count) were more expensive jobs. Likely pro or semi-pro jobs that were used for weddings and a christening. Sure is a chore to clean that stuff. I didn't want to use more filters; anything stronger, and it was starting to look like anime.
    Last edited by sanlyn; 21st Mar 2014 at 15:40.
    Quote Quote  



Similar Threads

Visit our sponsor! Try DVDFab and backup Blu-rays!