Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.

Messages - hsei

Pages: 1 ... 3 4 [5]
Bugs / New analysis mode
« on: September 24, 2010, 11:09:14 »
The new analysis mode is a nice feature, but has to be used with caution:
VBR files are rated totally false, the bitrate is taken from the first frame(s) where it is typically very low. Consequently the duration is calculated much too high (total size by first frame size).
Reading of further frames seems to take into account the bitrate in frame header, otherwise complete rubbish would appear.
I would advise to ether correct ratings for VBR to average bitrate or drop rating (at the moment)  for VBR at all.
In frequency analysis I found another problem: I got a rating of 100% for a file sampled at 44100 Hz with a max. frequency of 22050. Files should never contain frequencies up to the Nyquist frequency (half the sample rate), because that leads to annoying alias frequencies. That should lead to drastic reduction in quality rating. For the file in question I found no perceptible degradation, so the max. frequency value may be an analysis flaw.
Again: Be reluctant with analysis ratings, but with some caution it's a valuable feature. Calculating quality measures is a demanding task.

Wishlist / Re: CBR vs. VBR indication
« on: August 08, 2010, 10:23:13 »
I support getting as much information as feasible. In this case the duplicates differ mainly by size. It would be helpful to have at least an indication (e.g. background colour) of great deviations. Primary task of similarity is to identify almost identical files. Ranking duplicates with same content and different quality in a second step is a non-trivial task. Comparing bitrate, size etc. is not always reliable especially for reencoded material.

You better look at both similarity columns. Songs with standard scores below 80-85% typically have precise scores below 10%. But eventually there appear completely different songs with very high precise scores (nobody is perfect), but they show low standard scores and can be easily rejected by that.

Wishlist / Other Comparison Criteria
« on: July 13, 2010, 10:29:33 »
SNR feature could be extended:
Quality measure by frequency (band) content.
Rather often potentially high quality recordings (e.g. 192 KBit/s MP3) turn out to be just recompressions of e.g. 128 kBit/s material. Of course frequencies (or other details) lost at previous compressions can't be regained by that.

But besides that: This software is rather helpful since it does local comparison on a track to track basis and doesn't rely on a server database with entries of acoustic fingerprint vs. tags.

General / save
« on: July 11, 2010, 16:27:47 »
A possible workaround is not to shut down your computer but to switch to hibernate / power saving or whatever that is called in your OS. In Win7 it was even possible for me to pause scanning and to proceed from that point after restarting Windows.

Bugs / Performcence issues
« on: July 11, 2010, 12:00:17 »
My recommendation is to put more emphasis on *duration*. To look for music that starts the same but is different in length is maybe a nice feature but of no interest for 99% of users searching typically for real duplicates. By leaving that feature e.g. as an option and concentrating on files with approximatly the same duration would dramatically reduce search space and boost performance for larger collections. There's no additional cost since you compute track duration anyway and that's only done once per track -> O(N) and not per comparison -> O(N square).

Wishlist / Dont show duplicates under xx% similarity
« on: July 11, 2010, 11:33:51 »
No, it doesn't.
It only filters out groups with *all* entries below that threshold.
Maybe that's not perfect but it's better than nothing.

Bugs / "Copy All" in tag editor
« on: July 11, 2010, 11:10:06 »
The "apply" button only acts on the entry of the group which you selected at the beginning by "edit tags (CTRL-E)". The other entries you click thereafter can only be used as source of tag entries but not as target. This applies not only to "copy all" but also for "copying" single entries by clicking on the blue header above the field.
Try that.

Bugs / false 100% similarity
« on: July 11, 2010, 11:00:18 »
It's not only a problem of long durations: Having two files of e.g. 2 minutes with high similarity score and differing by 10 secs is a strong indication of corruption.
I actually use that for identifying corrupted files but at the moment it has to be done "manually" by looking for significant duration mismatches in high score groups.

Pages: 1 ... 3 4 [5]