Discussion in 'other anti-virus software' started by narenbisht, Jan 17, 2012.
Thanks for the test, I'm going to watch for this in the future.
Updated percentages, rounded to nearest %:
Comodo - 131/210 - 62%
Unthreat - 131/210 - 62%
Avira - 128/210 - 61%
AVG - 120/210 - 57%
Avast! - 104/210 - 50%
Foriclient Lite - 105/210 - 50%
Panda - 13/31 - 42%
MSE - 87/210 - 41%
Thanks for your time and effort narenbisht
Disable defense+ of comodo and test again.
That was with Def+ ? I'd hate to see its score without.
It says Comodo AV rather than CIS,so I'm thinking D+ wasn't used.If it was it'd unfairly skew the results,since the other products aren't full suites.
Perhaps Naren will confirm this?
I'm not very familiar with Unthreat,but it appears to be a competent newcomer.
To my knowledge Comodo AV does not come with defense+ . I think that packed with the firewall.
If it was with just Comodo's AV I've got to give them props for doing this well, especially against 0-day malware.
Are you testing signature scanning only? I mean, if the malware is available from a URL (-www.maliciousdomain.com/malware.exe-) are you testing the product during download or only its file-signature component (ie right-click scan)?
Some users of forum -safegroup.pl- lead tests of some AV (free and paid)...not all and not always the same AV...latest results of free AV
18 Jan. 2012
Avira 2012 - 123/162 (76%)
KPCD - 84/162 (51,9%)
16 Jan. 2012
MSE 2 - 89/152 (58,6%)
CIS 2012 - 115/152 (75,5 %)
UnThreat AV free - 119/152 (78,28%)
KPCD - 100/152 (65,8%)
MSE 2 - 75/124 (60,5%)
KPCD - 106/128 (82,8%)
Panda Cloud AV - 110/128 (86%)
AVG 2012 Free - 100/128 (78,13%)
CIS 2012 - 109/128 (85,2%)
KPCD - 106/128 (82,8%)
today (19. Jan.)
KPCD - 126/162 (77,77%)
Avira - 133/162 (82,1%)
D+ is disabled.
I am only testing signatures, heuristics, behaviour blocker, etc. Purely AV test.
This is just a simple RealTime Protection/Detection Test. I download 0-Day malware mainly from MDL, MBL, Malc0de & CleanMX & sometimes I look for other sites for samples. I collect 0-Day malware in a folder & check for the duplicate malware i.e same malware file. First RightClick Scan & remaining are executed to check if they are detected. Hope this is clear.
I think the testers are missing an essential part of Product evaluations. There is the assumption that whatever files are downloaded from the malware link sites are actual malware. This is far from the case! Depending on the day from 15-30% of these files are totally innocuous (keygens, cracks, etc).
This must be what is going on as the results that they are getting for some products (if indeed tested at default settings) are so far off from the multitude of previous results to make the current test suspect.
Updated Jan 19 2012, checkout 1st post.
This is the last & final update.
I have already mentioned not a professional test, Just a simple test.
And atleast Avira & Avast I also tested with All Threats Enabled in Avira & PUP enabled in Avast. The difference was not much, just 1-2 more detections.
I didn't added this detection in the results.
But as I said not much difference 1-2 more detection.
Thanks for your comparison of products. I'll take common everyday testing anytime . It means more to me than over controlled enviroments with what seems over controlled results. Test my product as long as it does well gets old. Along as everybody gets the same pitch lets see what happens.
Most of the abreviations I can figure out ichito but what is KPCD?
Despite being against posting policy the thread managed to stay alive until the end.
Kingsoft PC Doctor (KPD/KPCD)
Thanks for that.
Panda Cloud AV - 137/162 (84,6%)
Naren, can you please PM me the md5 checksums of the files you tested that were "not neutralized"? This would really help us improve the product for everyone else.
Sorry, I deleted the samples as I dont really keep them after testing as for any test I download the fresh samples again.
But one thing I have noticed, just go to any malware sites & download 30-40 malware samples & test against panda you will find few that are not neutralized.
The prob may be with panda's cleaning function. Is this prob with other Panda's products too or with PandaCloud only? If only with PandaCloud then why, especially when you have successful cleaning tools in your other products?
May be next time you test you will post also the MD5 otherwise it seems at bit all in the air....
Separate names with a comma.