AV Comparatives has released the newest report.

Yes because this is old samples.

Good effort by Panda and I feel that Avira will do very well in AV-C tests this year despite some concerns expressed by some members in another thread.

 

FPs aside, Avira performs consistently like before, no one is bothered with FP, its detection we are after and as usual Avira put all doubts to rest, Avast I am sorry needs to do some serious rethinking.

 

Osaban, I was going to add to my last sentence (but it was doing my head in, just estimating, no calculators needed lol), from looking at the report, Trustport would have been no.3 behind G-Data and Avira. Excellent detection rates, acceptable FPs, but slow as a snail scan speed though. Reason it might not be used on here, as reports of it being heavy. It might have improved though since people last spoke about it.

No.4 could be either PC Tools or Symantec (the latter with the fastest scan speed, similar detections to PC tools, only 2 more FPs). One might say PC Tools edged out Symantec on this one for almost a thousand more detections, but I'd give them a tie (PC Tools for more detections, two less FPs, but Symantec for 'blistering', yeah I'll use the word, 'blistering' scan speed ).

And if it weren't for the high number of false positives, then Panda (47 FPs) and McAfee (61 FPs) to close out the fourth and fifth spot. But these two need to improve....

So because of the false positives, and AV-C sees it as a concern, it'd be a close call between either F-Secure (97.8% detection / 2FPs) of ESET (97.7% detection / 3 FPs). ESET having slightly better scan speed. Give them a tie as well!

1. G-Data 2. Avira 3. Trustport 4*. PC Tools/Symantec 5**. F-Secure/ESET

*If FPs not a concern, Panda
** If FPs not a concern, McAfee

 

Serious rethinking after scoring Advanced+ ? I'm sure some heads are going to roll.

 

I expected it to surpass Avira 9, with all the potential it has, I am sure thats possible. However so far, Avast's improvement is nominal over its previous version so yes, some serious introspection allright.

 

Ummm... AFAIK, Blackcat has been testing TM's (Trend Micro's) current beta. As I recall, he says that TM-beta's detection is MUCH better than TM's current version (the one tested). I hope so. Even though I am not a TM fan, we need plenty of top-tier AVs, & I would like to see TM join that group.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~

As to K7...
K7 is a very small AV outfit in India. I bought a K7 license a couple of years back -- just to contribute to a small & energetic outfit.

I used K7 real-time off & on for about a year, then let the license expire. At that time, I found K7 to be incredibly light, & was impressed by the fact that it updated its sigs 1 or more times every day. K7 never blipped, of course, because I almost never venture into shark-infested waters.

I applaud K7's courage in subjecting itself to AV-C's testing, especially considering the fact that participation in AV-C's tests costs money. Although K7 is not yet a top-tier AV, their willingness to participate in AV-C's rigorous testing gives evidence that they intend to get better.

I wish them well, & might buy another license one of these days when I have some loose change in my pocket.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

As to Avira...
Yowzah!

 

K7 did quite well in detection, and scan speed isn't so bad. The 193 FPs let it down, otherwise it would have been awarded Advanced.

 

IMO, the whole summary of this report is that new contenders are really showing their muscles...Panda did a very good job and same with PC-Tools and K-7 ...Only 193 fp's let it down...But it is far more better then Trend Micro and AVG in terms of detection.

As far as McAfee is concerned they are really doing their tough job and now it shows...I am slightly disappointed with Kaspersky and Avast..because of their slow improvement...I found them at the same place where they left last year..

 

I'm pretty sure that PC Tools uses Norton's engine. The numbers are quite similar.

 

Another TOTALLY USELESS on-demand test. These tests tell you absolutely nothing about the overall capabilities of the product.

Andreas Clementi, you need to stop doing these useless tests.

 

Yeah me too think so...But one interesting thing i have noted that without an internet connection the detection rate of PANDA, McAfee and Trend Micro drastically fell down.

Without Internet Connection:-

Panda Scored 73.3%
McAfee Scored 94.9%
Trend Micro 68.5%

This shows that all these AV's does not contain that much of signatures of malwares in their offline database...They totally depend on their in-the-cloud technology..

 

+1

Is there any reason to stay offline nowadays?

 

Fair call. But instead of launching files, I know I just right-clicked and scanned 2 x 500 GB external drives. One from work, and another that was mine and a friend was using. I think many people still see the benefit in these tests.

 

No..but still in many parts of my country people do stay offline...so it matters for them...

 

I believe lot's will do worse without internet connection.
Lot's of vendors are trusting these days on their "cloud"

 

I know that they are trusting lot on their cloud technology, but still i believe that it would be better if they add each and every signature in their database for those ppl who cannot be online all the time..

 

An on-demand AV can be useful in several scenarios. Here are some examples that come to mind:

1- If one is using DW (DefenseWall) and downloads an exe, DW always categorizes it as "untrusted." To install that exe in "untrusted" status is usually futile.

1a- Thus, that exe must needs be "run as trusted" or else have its status changed to trusted.

1b- At such times it is handy to have an AV (&/or MBAM, et alia) do an on-demand check of that exe.

2- When checking out an exe via sandbox, I have heard that some malware nowadays is able to detect when it is being run "virtualized" &, hence, will "play nice" until it gets installed & executed for real. Here again an on-demand app *might* spot the nastiness of such an exe.

3- Also, if one helps a neophyte friend install his/her first security apps, it is handy to have a powerful on-demand app do a full scan of that person's computer.

4- A couple of my friends surf without running ANY security apps in real-time. They make much use of full & frequent on-demand scans, as well as doing frequent system disk imaging.

5- Further, more & more AVs are incorporating better emulators/heuristics, as well as behavior blockers, either or both of which capabilities can kick in during on-demand scans. Thus, on-demand scans are becoming much more than the simple application of blacklist sigs.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Besides the examples listed, there are several other good uses for on-demand checks. Ergo, I am fully persuaded that AV-C's on-demand tests provide useful insight concerning one significant aspect of an AV's protective potential.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

I agree BUT -- a primary source of infections is the internet. It's not impossible to be infected while offline, but it is rather difficult to do so.

It is true that cloud-based AVs are mostly offline when one's computer is offline, but so, also, are most of the malware threats mostly offline.

To become infected apart from the internet usually calls for some degree of ignorance or folly to be exercised.

In other words, a person must usually have his brain offline in order to get his computer infected offline.

 

Don't forget that Panda doesn't use their Cloud for on-access (real-time scans) unless in CloudAntivirus. This will make Panda good scanner but not that good protector with 70% without the Cloud.

These tests are still not that important because they don't consider HIPS or other protection .

I wonder why AV-Comparatives still uses Windows XP for their tests ? Why not Windows 7 ?

 

You are right..that primary source of infection is the internet, but imagine if some malware get inside your system and tamper your host file and other things which results in network disconnection...which lead your system offline..then these so called AV's who rely on their cloud tech. will get crippled...

By the way, see some Tele-Commercials from K7 Computing:-

http://www.k7computing.in/telecommercials

 

This is not true. Panda AV/IS on-access also use cloud-scanning, just a slightly different implementation.

 

completely agree, and people know this, but all are like little kids, waiting for christmas, the results come and they care not about the test, just the results.

these sort of tests give ZERO indication about their detection for the customer, so why bother post the results?

im sure IBK has been getting training sessions from M.R.G, he needs to stop wasting time doing this tripe, and get on doing some proper dynamic testing, even after AVC themselfs mention the advantages of proper dynamic testing, they go back and do a completely useless one again, one that i thought was completely laughed-off by the major of people with brains.

why bother wasting resources and time on producing these results, when they cannot be trusted?

sorry AVC but you are poor testers, i shall wait for the next MARX one.

and before all those 'type' of people reply, dont bother - you either like my comments, hate them even though you know they are true, or will enter the thread with the 'reply' button to start some hassle about your 'beloved tester IBK at AVC', either way, thats for you to do, i shant be wasting my time to entertain these fanboys.

 

I am really satisfied the result of ESET, and my opinion is that 4.2 really made is an improvement over the tested version.
Good detection rates, and not even on ESET's speciality but on-demand, extremely low FP's and a rock solid scan speed.

GData shows they are really doing good, I just still find it too heavy.
MSE, great performance, because they are just like ESET more focused on proactive than on on-demand.
Also McAfee surprised me, they are actually really improving. And a good job by Panda.
Little dissappointed in Kaspersky, I always thought ESET and Kaspersky were (overall, all tests, last 5 years) they only av's I would consider paying for.

 

I am sure that this "slightly different implementation" will result you in a failure in the Dynamic tests . But we'll see , right ?

 

I disagree that on-demand scan test can be useful and I kind of agree with Zombini that AVC should immediately stop doing such "outdated by type" on-demand tests. They have done some very good dynamic at the end of 2009. What a percent means without executing the file . We need to see that the antivirus does ACTIVE PROTECTION . On-demand scans compare nothing but the blacklists of all candidates.
We have a porter who has a very big blacklists and does nothing but stops those who are in this list.Everybody else not on the list is allowed to get in the house.

The Dynamic test is the interesting one where not only the porter but ALL guards interact (take part in the big game)

 

Wait some months for dynamic test results.